Disclosed herein is an image processing apparatus including a color-flicker detection processing section configured to detect generation of color flickers from a plurality of images taken at different exposure times, wherein the color-flicker detection processing section: acquires a ratio of any specific one of a plurality of color signals of the taken images to another one of the color signals; obtains an evaluation value of the color flickers from the acquired ratio; and determines whether or not the color flickers have been generated on the basis of the magnitude of the evaluation value.

Patent
   8670043
Priority
Jun 24 2010
Filed
Jun 07 2011
Issued
Mar 11 2014
Expiry
Jan 23 2032
Extension
230 days
Assg.orig
Entity
Large
3
6
EXPIRED
8. An image processing method, comprising:
detecting a color flicker from a plurality of images, where a first image of said plurality of images is acquired at a different exposure time than a second image of said plurality of images, wherein said color flicker is detected based on:
acquiring a ratio for each of said plurality of images between any specific one of a plurality of color signals of said each of said plurality of images to another one of said plurality of color signals;
obtaining an evaluation value from a comparison of said ratio corresponding to said first image and said ratio corresponding to said second image;
computing a degree of certainty value based on determining that said color flicker has been generated based on a magnitude of said evaluation value; and
performing white-balance adjustment processing by multiplying said plurality of images or a synthesized image obtained as a result of synthesizing said plurality of images by white-balance gains, wherein said white-balance gains include a gain generated from blending gains for said plurality of images based on a blending ratio obtained from said degree of certainty.
1. An image processing apparatus, comprising:
a color-flicker detection processing section configured to detect a color flicker from a plurality of images, where a first image of said plurality of images is acquired at a different exposure time than a second image of said plurality of images, wherein the color-flicker detection processing section is configured to:
acquire an integrated value of each of a plurality of color signals for each of said plurality of images, wherein said integrated value is acquired on a range obtained by adding a noise margin to an area of said each of said plurality of images, wherein said area corresponds to at least one pixel that outputs a value when said each of said plurality of images was acquired;
utilize said integrated value to acquire a ratio for said each of said plurality of images between any specific one of said plurality of color signals of said each of said plurality of images to another one of said plurality of color signals;
obtain an evaluation value from a comparison of said ratio corresponding to said first image and said ratio corresponding to said second image; and
determine whether or not said color flicker has been generated based on a magnitude of said evaluation value.
9. A non-transitory computer-readable medium storing program code executable by a processing unit to perform operation comprising:
detecting a color flicker from a plurality of images, where a first image of said plurality of images is acquired at a different exposure time than a second image of said plurality of images, wherein said color flicker is detected based on:
acquiring a ratio for each of said plurality of images between any specific one of a plurality of color signals of said each of said plurality of images to another one of said plurality of color signals;
obtaining an evaluation value from a comparison of said ratio corresponding to said first image and said ratio corresponding to said second image;
computing a degree of certainty value based on determining that said color flicker has been generated based on a magnitude of said evaluation value; and
performing white-balance adjustment processing by multiplying said plurality of images or a synthesized image obtained as a result of synthesizing said plurality of images by white-balance gains, wherein said white-balance gains include a gain generated from blending gains for said plurality of images based on a blending ratio obtained from said degree of certainty.
10. An image processing apparatus, comprising:
a color-flicker detection processing section configured to detect a color flicker from a plurality of images, wherein a first image of said plurality of images is acquired at a different exposure time than a second image of said plurality of images, wherein the color-flicker detection processing section is configured to:
acquire a ratio for each of said plurality of images between any specific one of a plurality of color signals of said each of said plurality of images to another one of said plurality of color signals;
obtain an evaluation value from a comparison of said ratio corresponding to said first image and said ratio corresponding to said second image; and
compute a degree of certainty value when said color-flicker detection processing section determines that said color flicker has been generated; and
a white-balance adjustment processing section configured to perform white-balance adjustment processing by multiplying said plurality of images or a synthesized image obtained as a result of synthesizing said plurality of images by white-balance gains, wherein said white-balance gains include a gain generated from blending gains for said plurality of images based on a blending ratio obtained from said degree of certainty.
2. The image processing apparatus according to claim 1, wherein said color-flicker detection processing section is configured to:
determine that said color flicker has not been generated if said evaluation value is within a predetermined dead zone; or
determine that said color flicker has been generated if said evaluation value is outside said predetermined dead zone.
3. The image processing apparatus according to claim 1, wherein said color-flicker detection processing section is configured to:
compute a degree of certainty value when said color-flicker detection processing section determines that said color flicker has been generated.
4. The image processing apparatus according to claim 3, further comprising:
a white-balance adjustment processing section configured to perform a white-balance adjustment processing by multiplying said plurality of images or a synthesized image obtained as a result of synthesizing said plurality of images by white-balance gains,
wherein said white-balance gains include a gain generated from blending gains for said plurality of images based on a blending ratio obtained from said degree of certainty.
5. The image processing apparatus according to claim 1, wherein said color-flicker detection processing section comprises a movement determination section configured to carry out movement determination, and is further configured to detect said color flicker by making use of a result of said movement determination.
6. The image processing apparatus according to claim 1, wherein said color-flicker detection processing section is further configured to:
divide an input image into a plurality of small areas; and
detect said color flicker for each of said plurality of small areas.
7. The image taking apparatus according to claim 1, further comprising:
an image taking device for acquiring each of said plurality of images at different exposure times.

The present disclosure relates to an image processing apparatus for producing an image with a wide dynamic range by making use of a plurality of images taken at different exposure times, an image taking apparatus employing the image processing apparatus, an image processing method for the image processing apparatus and an image processing program implementing the image processing method.

More particularly, the present disclosure relates to a mechanism for suppressing flickers generated in a video due to frequency variations included in illuminated light in an operation to generate an image of an image taking object by making use of a variety of image taking apparatus. These frequency variations are caused by power-supply frequencies. The flickers are also referred to as optical-source flickers.

An image taking apparatus for taking an image of an image taking object includes a mechanism for controlling the quantity of light incident to a pixel section of the apparatus. The image taking apparatus is typically a solid-state image taking apparatus of a CCD (Charge Couple Device), MOS (Metal Oxide Semiconductor) or CMOS (Complementary MOS) type. In the following description, the pixel section is also referred to as an image taking section, a pixel array section, an image sensor or an opto-electrical conversion sensor whereas the mechanism is referred to as an incident-light quantity controller.

The incident-light quantity controller is a controller having a mechanical diaphragm mechanism provided on an image taking lens thereof or a controller having a mechanical shutter mechanism provided on an image taking lens thereof. In the following description, the mechanical diaphragm mechanism is referred to as a mechanical iris whereas the mechanical shutter mechanism is referred to merely as a mechanical shutter. As an alternative, the incident-light quantity controller can also be a controller having the so-called electronic shutter function capable of controlling the length of an accumulation time of signal electrical charge in the pixel section of the solid-state image taking apparatus. In the following description, the accumulation time is also referred to as an exposure time.

The mechanical iris and the mechanical shutter can be used independently of each other. However, the mechanical iris can also be used by combining the mechanical iris with the mechanical shutter or the electronic shutter.

By the way, a variety of image taking apparatus do not raise a problem if the apparatus are used for taking an image by making use of an optical source put in a steady state in which the brightness of light generated by the source does not change. If an image is taken by making use of an optical source such as a fluorescent lamp having a periodical light emitting characteristic and operating asynchronously with the exposure period of the semiconductor image taking apparatus, however, optical source flickers are generated.

It is to be noted that the optical source flickers are distinguished from luminance flickers which are flickers of screen luminance and from color reproduction flickers which are also referred to simply as color flickers or color rolling.

The optical source flickers are perceived as a phenomenon in which a video signal changes due to a relation between luminance changes of an optical source and the exposure period of an image taking apparatus.

For example, the luminance signal component of a video signal changes due to luminance changes of an optical source making use of a commercial power supply having a frequency f and due to beat components having a field period fv of the image taking apparatus. In this case, the luminance changes are changes within a period of 1/nf seconds where reference symbol n is normally the integer 2. Since the luminance signal component of a video signal changes, the output video also varies as well at a period also related to the afterglow characteristic of the eye of a human being. A phenomenon in which image flickers are felt is referred to as luminance flickers.

For example, luminance flickers are generated with ease in an area in which the NTSC system having a field frequency of 60 Hz is adopted and the frequency f of the commercial power supply is 50 Hz. Luminance flickers are also generated with ease in an area in which the PAL system having a field frequency of 50 Hz is adopted and the frequency f of the commercial power supply is 60 Hz. In addition, in comparison with the electrical light bulb, the luminance of the fluorescent lamp changes due to the light emitting characteristic of the fluorescent lamp so that luminance flickers are generated very considerably by the fluorescent lamp.

It is to be noted that a statement saying a field frequency of 60 Hz can be said in other words as a statement saying a frame frequency of 30 fps. Speaking more accurately, the field frequency is 59.94 Hz. On the other hand, a statement saying a field frequency of 50 Hz can be said in other words as a statement saying a frame frequency of 25 fps.

For example, the emission period of the fluorescent lamp is 10 ms whereas the period of the exposure operation in the NTSC system having a field frequency of 60 Hz is 16.7 ms. In this case, the lowest common multiple of the emission period of the fluorescent lamp and the period of the exposure operation in the NTSC system is 50 ms. That is to say, in three exposure operations, the relation between the emission period of the fluorescent lamp and the period of the exposure operation in the NTSC system is restored. Thus, there are three kinds of exposure period. Differences of the levels of signals output by the solid-state image taking apparatus between these three exposure periods cause flickers to be generated at a flicker frequency F of 20 Hz.

In addition, if the function of the electronic shutter is used, the higher the speed of the shutter operating in the shutter mode, the shorter the time included in the one field period as an accumulation time for accumulating electric charge in the solid-state image taking apparatus.

Thus, the amplitude of flickers becomes larger than that for a normal shutter speed of 1/60 seconds. The higher the speed of the electronic shutter, the more striking the generated flickers. As a result, flickers including mainly image luminance flickers appear on the screen, causing the quality of the image displayed on the screen to deteriorate considerably.

In addition, the green, red and blue colors are three colors of a fluorescent substance used in a fluorescent lamp. Even though the emissions of the three colors start with the same timing, the light quantities of the three colors decrease at different rates so that the three colors disappear eventually at different times. That is to say, the light emitted by the fluorescent lamp changes its spectrum with the lapse of time.

In general, the emission time of the green color is particularly longest among the three colors. On the other hand, the emission time of the blue color is shortest among the three colors. That is to say, the emission time of the red color is between those of the green and blue colors.

Thus, depending on the shutter timing of the shutter having a high speed, only one or two color components of the emitted light can be taken in some cases.

In an operation to take an image by making use of an electronic shutter having a high speed, a difference in taken spectrum appears a color change. As described above, color reproduction flickers are also referred to as color flickers or color rolling.

In particular, the light component of the blue color cannot be taken as color flickers. In many cases, if an image taking operation is carried out normally, the light component of the blue color is taken inadvertently as the light component of the yellow color.

By the way, Japanese Patent Laid-open No. Hei 7-322147 describes a solid-state image taking apparatus for generating a video signal with a wide dynamic range by synthesizing images having different exposure quantities.

In the case of this image taking apparatus, if the image of a bright portion is taken at a normal image-taking exposure time in order to obtain a wide dynamic range, the output has a value exceeding the saturation level of an opto-electrical conversion sensor employed in the image taking apparatus so that the value cannot be used. Thus, it is necessary to take images each having a small exposure quantity and then synthesize the taken images.

FIG. 1 is a conceptual diagram showing relations between the sensor output and the light quantity for images required for generating an image having a wide dynamic range.

Images taken at a normal exposure time include an image that cannot be acquired due to the fact that the light quantity of this unacquirable image exceeds the saturation level of the opto-electrical conversion sensor. For the image that cannot be acquired, the exposure time is shortened in order to prevent the light quantity from exceeding the saturation level of the opto-electrical conversion sensor so that meaningful information can be obtained.

The value of each pixel of the image taken at a short exposure time is multiplied by the reciprocal of an exposure ratio described below or added to an offset in order to produce a resulting pixel value. Then, for every pixel, the resulting pixel value is compared with the value of a corresponding pixel of an image taken at a long exposure time in order to select a particular one of the two pixels compared with each other and discard the other one. In this way, a synthesized image having a wide dynamic range can be generated.

It is to be noted that the exposure ratio is a ratio of the shutter speed of a normally taken image to the shutter speed of an image taken at a short exposure time.

As described above, in order to generate an image having a wide dynamic range, it is necessary to reduce the exposure time to a relatively short one by a shutter speed corresponding to an exposure time of (1/several thousands) seconds. In the following description, the exposure time is also referred to as a shutter time. Thus, an image taken at such a short exposure time has generated color flickers. In the following description, the image taken at such a short exposure time is referred to as a short exposure-time image.

On the other hand, an image taken at an exposure time equal to or longer than that for a standard video signal is normally obtained at a shutter speed corresponding to an exposure time of typically 1/60 or 1/100 seconds in many cases. In the following description, the image taken at such a long exposure time is referred to as a long exposure-time image. Thus, as explained before, a long exposure-time image is hardly affected by color flickers.

Accordingly, the effect of color flickers on a short exposure-time image is big but the effect of color flickers on a long exposure-time image is small. Thus, as a result, on a portion of a synthesized image having a wide dynamic range, color flickers are generated. Accordingly, since the synthesized image is different from the image actually seen by the eyes of the user, the user inevitably feels the synthesized image in an unnatural way.

In order to solve the problem described above, there has been made a proposal to carry out white-balance adjustment processing for an image taken by making use of short exposure time light and white-balance adjustment processing for an image taken by making use of long exposure time light separately from each other.

FIG. 2 is a block diagram showing a first typical image taking apparatus 1 for carrying out white-balance adjustment processing for an image taken by making use of short exposure time light and white-balance adjustment processing for an image taken by making use of long exposure time light separately from each other.

As shown in the figure, the image taking apparatus 1 employs a solid-state image taking device 2, front-stage processing sections 3-1 and 3-2, conversion processing sections 4-1 and 4-2 each used for converting RAW data into RGB data and an integrated-value computation processing section 5.

The image taking apparatus 1 also has white-balance adjustment processing sections 6-1 and 6-2, amplifiers 7-1 and 7-2, a synthesizing processing section 8, a rear-stage processing section 9 and a CPU 10.

The integrated-value computation processing section 5 employed in the image taking apparatus 1 computes an integrated value of the RGB data of the long exposure time light and an integrated value of the RGB data of the short exposure time light. The integrated-value computation processing section 5 supplies the computed integrated values to the CPU 10. On the basis of the integrated values, the CPU 10 computes a white-balance gain to be used as a multiplier for the long exposure-time image and a white-balance gain to be used as a multiplier for the short exposure-time image.

The CPU 10 supplies the computed white-balance gains to the white-balance adjustment processing sections 6-1 and 6-2. The white-balance adjustment processing section 6-1 carries out white-balance adjustment processing to multiply the RGB data of the long exposure-time image by the white-balance gain for the data whereas the white-balance adjustment processing section 6-2 carries out white-balance adjustment processing to multiply the RGB data of the short exposure-time image by the white-balance gain for the data.

As described above, by adoption of this system configuration, it is possible to multiply the long exposure-time image and the short exposure-time image by their respective white-balance gains different from each other.

FIG. 3 is a block diagram showing a second typical image taking apparatus 1A for carrying out white-balance adjustment processing for an image taken by making use of short exposure time light and white-balance adjustment processing for an image taken by making use of long exposure time light separately from each other.

The image taking apparatus 1A shown in FIG. 3 is a typical modification of the image taking apparatus 1 shown in FIG. 2.

In the image taking apparatus 1A shown in FIG. 3, the integrated-value computation processing section 5 and the white-balance adjustment processing section 6 are provided at stages following the synthesizing processing section 8.

If it is possible to supply information indicating the long exposure-time image or the short exposure-time image from which a pixel has been selected in the synthesizing processing carried out by the synthesizing processing section 8 of the image taking apparatus 1A shown in FIG. 3 as a flag FLG for every pixel to the following stage, the same processing as that of the image taking apparatus 1 shown in FIG. 2 can be carried out in the post-synthesizing processing. As described earlier, the processing carried out by the image taking apparatus 1 shown in FIG. 2 includes the integrated-value computation processing and the white-balance adjustment processing.

The configuration shown in FIG. 3 offers a merit that the number of pixels subjected to the integrated-value computation processing and the white-balance adjustment processing is half the number of pixels subjected to the integrated-value computation processing and the white-balance adjustment processing which are carried out in the configuration shown in FIG. 2. That is to say, the image taking apparatus 1 shown in FIG. 2 carries out the processing on pixels of two images, i.e., the long exposure-time image and the short exposure-time image, whereas the image taking apparatus 1A shown in FIG. 3 carries out the processing on only pixels of one image, i.e., the synthesized image. Thus, the image taking apparatus 1A shown in FIG. 3 has a circuit size smaller than the image taking apparatus 1 shown in FIG. 2.

By adoption of the system configuration described above by referring to FIG. 2 or 3, it is possible to make use of white-balance gains as multipliers for the long exposure-time image and white-balance gains, which are different from the white-balance gains for the long exposure-time image, as multipliers for the short exposure-time image.

If integration processing is carried out on each of the long exposure-time image and the short exposure-time image, however, the number of pixels valid for the integration processing of the images is small. A valid pixel is a pixel having a value not exceeding the maximum sensor output and a pixel not having a value of 0 or not having an extremely small value.

Thus, in some cases, correction is undesirably carried out as correction lured by things other than color flickers caused by optical-source variations to serve as an object of suppression. The things other than color flickers include the color of a moving object.

Even in a normal case not generating color flickers caused by an optical source, white-balance adjustment processing carried out mistakenly on a portion of a synthesized image inadvertently gives a color to the portion.

In addition, in order to solve the problems described above, it is necessary to carry out processing to impose a commonly known restriction and/or processing to determine whether or not an optical source is a flicker optical source like one described in Japanese Patent Laid-open No. Hei 11-75109. It is quite within the bounds of possibility that the processing to impose the commonly known restriction is carried out if the image taking operation is performed at a high shutter speed.

In the present state of the art, however, it is difficult to distinguish incorrect detection of a moving object as described earlier from detection of actually generated color flickers with a high degree of precision.

In order to solve the problem described above, there is provided an image processing apparatus capable of detecting color rolling in a more robust way and suppressing the detected color rolling, an image taking apparatus employing the image processing apparatus, an image processing method adopted by the image processing apparatus as well as an image processing program implementing the image processing method.

According to a first mode of the present disclosure, there is provided an image processing apparatus including a color-flicker detection processing section configured to detect generation of color flickers from a plurality of images taken at different exposure times, wherein the color-flicker detection processing section: acquires a ratio of any specific one of a plurality of color signals of the taken images to another one of the color signals; obtains an evaluation value of the color flickers from the acquired ratio; and determines whether or not the color flickers have been generated on the basis of the magnitude of the evaluation value.

According to a second mode of the present disclosure, there is provided an image taking apparatus including: an image taking device for taking a plurality of images at different exposure times; and a color-flicker detection processing section configured to detect generation of color flickers from the images taken at different exposure times wherein the color-flicker detection processing section: acquires a ratio of any specific one of a plurality of color signals of the taken images to another one of the color signals; obtains an evaluation value of the color flickers from the acquired ratio; and determines whether or not the color flickers have been generated on the basis of the magnitude of the evaluation value.

According to a third mode of the present disclosure, there is provided an image processing method including detecting generation of color flickers from a plurality of images taken at different exposure times, the method further including: acquiring a ratio of any specific one of a plurality of color signals of the taken images to another one of the color signals; obtaining an evaluation value of the color flickers from the acquired ratio; and determining whether or not the color flickers have been generated on the basis of the magnitude of the evaluation value.

According to a fourth mode of the present disclosure, there is provided an image processing program to be executed by a computer for carrying out image processing including color-flicker detection processing of detecting generation of color flickers from a plurality of images, which are taken at different exposure times, by execution of: acquiring a ratio of any specific one of a plurality of color signals of the taken images to another one of the color signals; obtaining an evaluation value of the color flickers from the acquired ratio; and determining whether or not the color flickers have been generated on the basis of the magnitude of the evaluation value.

In accordance with the present disclosure, it is possible to detect color rolling in a more robust way and suppress the detected color rolling.

FIG. 1 is a conceptual diagram showing relations between the sensor output and the light quantity for images required for generating an image having a wide dynamic range;

FIG. 2 is a block diagram showing a first typical image taking apparatus for carrying out white-balance adjustment processing for an image taken by making use of short exposure time light and white-balance adjustment processing for an image taken by making use of long exposure time light separately from each other;

FIG. 3 is a block diagram showing a second typical image taking apparatus for carrying out white-balance adjustment processing for an image taken by making use of short exposure time light and white-balance adjustment processing for an image taken by making use of long exposure time light separately from each other;

FIG. 4 serves as a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a first embodiment of the present disclosure as well as a block diagram showing a typical configuration in which white-balance adjustment processing is carried out at the rear stage;

FIG. 5 shows a flowchart representing processing to detect color flickers in accordance with the first embodiment;

FIG. 6 shows a flowchart representing processing to integrate R, G and B pixel values;

FIG. 7 is a conceptual diagram to be referred to in description of a method for specifying a range of integration of pixel values for a case in which a noise margin is not taken into consideration;

FIG. 8 is a conceptual diagram to be referred to in description of a method for specifying a range of integration of pixel values for a case in which a noise margin is taken into consideration;

FIG. 9 is a conceptual diagram showing a ratio of R to G and a ratio of B to G for a short exposure-time image;

FIG. 10 is a conceptual diagram showing a ratio of R to G and a ratio of B to G for a long exposure-time image;

FIG. 11 is a diagram showing typical color-flicker evaluation values obtained by making use of equations;

FIG. 12 is an explanatory diagram to be referred to in description of a typical color-flicker determination method;

FIG. 13 is a block diagram showing a typical configuration of a color-flicker detection processing section;

FIG. 14 shows a flowchart representing color-flicker detection and certainty-degree computation processing;

FIG. 15 is a diagram to be referred to in explanation of a method for computing the degree of certainty at which color flickers are detected;

FIG. 16 is a diagram showing degrees of certainty at which color flickers are detected;

FIG. 17 is a diagram to be referred to in explanation of another method for computing the degree of certainty at which color flickers are detected;

FIG. 18 is a diagram showing a white-balance gain computation system for computing different final white-balance gains for short and long exposure-time images;

FIG. 19 is an explanatory diagram to be referred to in description of a method for computing a blending ratio α;

FIG. 20 is a diagram showing another white-balance gain computation system for computing different final white-balance gains for short and long exposure-time images;

FIG. 21 is a block diagram showing a typical configuration of a color-flicker detection processing section having a function to suppress color flickers;

FIG. 22 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a second embodiment of the present disclosure;

FIG. 23 shows a flowchart representing processing to detect color flickers in accordance with the second embodiment;

FIG. 24 is a block diagram showing a typical configuration of a color-flicker detection processing section according to the second embodiment;

FIG. 25 shows a flowchart representing processing for a case in which static/dynamic-state determination for R, G and B integrated values is included as an additional condition in accordance with a third embodiment of the present disclosure;

FIG. 26 is a block diagram showing a typical configuration of a color-flicker detection processing section provided with a static/dynamic-state determination function in accordance with the third embodiment;

FIG. 27 shows a flowchart representing processing to integrate evaluation values in accordance with a fourth embodiment of the present disclosure instead of integrating R, G and B integrated values;

FIG. 28 shows a flowchart representing color-flicker determination processing for a case in which the processing to integrate evaluation values has been carried out;

FIG. 29 serves as an explanatory diagram to be referred to in description of a case in which processing to divide a color-flicker detection area is carried out in accordance with a fifth embodiment of the present disclosure as well as a diagram showing a typical situation in which a plurality of optical sources exist;

FIG. 30 is an explanatory diagram to be referred to in description of a case in which processing to divide a color-flicker detection area is carried out in accordance with the fifth embodiment;

FIG. 31 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a sixth embodiment of the present disclosure;

FIG. 32 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a seventh embodiment of the present disclosure;

FIG. 33 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to an eighth embodiment of the present disclosure;

FIG. 34 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a ninth embodiment of the present disclosure;

FIG. 35 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a tenth embodiment of the present disclosure;

FIG. 36 is a block diagram showing a typical configuration of a synthesizing processing section adopting a synthesizing method for suppressing a false color generated in a blend area in accordance with the tenth embodiment;

FIG. 37 shows a flowchart representing processing to determine the value of a long/short exposure-time select flag in accordance with the tenth embodiment;

FIGS. 38A and 38B are a plurality of conceptual diagrams to be referred to in explanation of a reason why a false color generated in a blend area can be suppressed in accordance with the tenth embodiment;

FIG. 39 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to an eleventh embodiment of the present disclosure;

FIG. 40 is a diagram showing a typical configuration of a synthesizing processing section applying a synthesizing method to taken images in order to suppress a false color generated in a blend area in accordance with the eleventh embodiment;

FIG. 41 is a diagram showing a typical configuration of a white-balance adjustment processing section applying a synthesizing method to white-balance gains in order to suppress a false color generated in a blend area in accordance with the eleventh embodiment;

FIGS. 42A and 42B are a plurality of conceptual diagrams to be referred to in explanation of a reason why a false color generated in a blend area can be suppressed by making use of the synthesizing processing section and the white-balance adjustment processing section which are provided in accordance with the eleventh embodiment;

FIG. 43 shows a flowchart representing processing to compute a blending ratio in accordance with a twelfth embodiment of the present disclosure;

FIG. 44 is a block diagram showing a typical configuration of a synthesizing processing section provided with a static/dynamic-state determination function according to the twelfth embodiment to serve as a modification of the synthesizing processing section shown in FIG. 36;

FIG. 45 is a block diagram showing a typical configuration of a synthesizing processing section provided with a static/dynamic-state determination function according to the twelfth embodiment to serve as a modification of the synthesizing processing section shown in FIG. 40;

FIG. 46 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a thirteenth embodiment of the present disclosure;

FIG. 47 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a fourteenth embodiment of the present disclosure; and

FIG. 48 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to a fifteenth embodiment of the present disclosure.

Embodiments of the present disclosure are explained by referring to the diagrams.

It is to be noted that the embodiments are described in chapters arranged as follows:

1: First Embodiment Implementing a First Typical Image Processing Apparatus of an Image Taking Apparatus

2: Second Embodiment Implementing a Second Typical Image Processing Apparatus of an Image Taking Apparatus

3: Third Embodiment Implementing a Third Typical Image Processing Apparatus of an Image Taking Apparatus

4: Fourth Embodiment Implementing a Fourth Typical Image Processing Apparatus of an Image Taking Apparatus

5: Fifth Embodiment Implementing a Fifth Typical Image Processing Apparatus of an Image Taking Apparatus

6: Sixth Embodiment Implementing a Sixth Typical Image Processing Apparatus of an Image Taking Apparatus

7: Seventh Embodiment Implementing a Seventh Typical Image Processing Apparatus of an Image Taking Apparatus

8: Eighth Embodiment Implementing an Eighth Typical Image Processing Apparatus of an Image Taking Apparatus

9: Ninth Embodiment Implementing a Ninth Typical Image Processing Apparatus of an Image Taking Apparatus

10: Tenth Embodiment Implementing a Tenth Typical Image Processing Apparatus of an Image Taking Apparatus

11: Eleventh Embodiment Implementing an Eleventh Typical Image Processing Apparatus of an Image Taking Apparatus

12: Twelfth Embodiment Implementing a Twelfth Typical Image Processing Apparatus of an Image Taking Apparatus

13: Thirteenth Embodiment Implementing a Thirteenth Typical Image Processing Apparatus of an Image Taking Apparatus

14: Fourteenth Embodiment Implementing a Fourteenth Typical Image Processing Apparatus of an Image Taking Apparatus

15: Fifteenth Embodiment Implementing a Fifteenth Typical Image Processing Apparatus of an Image Taking Apparatus

FIG. 4 is a block diagram showing a typical configuration of an image taking apparatus 100 employing an image processing apparatus 120 according to a first embodiment of the present disclosure.

Configuration of the Image Taking Apparatus

The image taking apparatus 100 employs a solid-state image taking device 110 and the image processing apparatus 120.

The image processing apparatus 120 has front-stage processing sections 121 and 122, amplifiers 123 and 124 as well as a color-flicker detection processing section 125.

In addition, the image processing apparatus 120 also includes a synthesizing processing section 126, an RAW-RGB conversion processing section 127, an integrated-value computation processing section 128, a white-balance adjustment processing section 129, a rear-stage processing section 130 and a CPU (central processing unit) 131 functioning as a control section.

The blocks composing the image taking apparatus 100 are explained in an order according to the sequence of processing operations carried out to generate an image having a wide dynamic range.

The solid-state image taking device 110 is configured as an image sensor such as a CCD image sensor or a CMOS image sensor. In this specification of the present disclosure, the technical term ‘sensor’ is also used to imply the solid-state image taking device 110.

The solid-state image taking device 110 carries out opto-electrical conversion to convert a taken image created by an optical system not shown in the figure into a digital signal, supplying the digital signal to the front-stage processing sections 121 and 122 employed in the image processing apparatus 120.

In the solid-state image taking device 110, a plurality of images taken at different exposure times are created in order to generate an image having a wide dynamic range. To put it concretely, the solid-state image taking device 110 outputs digital signals representing at least two images taken at different exposure times.

In the configuration shown in the figure, reference notation Short denotes a short exposure-time image which is an image taken at a short exposure time. On the other hand, reference notation Long denotes a long exposure-time image which is an image taken at a long exposure time equal to a normal exposure time longer than the short exposure time or even longer than the normal exposure time. The short exposure-time image and the long exposure-time image are used for generating an image having a wide dynamic range.

The front-stage processing carried out by the front-stage processing sections 121 and 122 is processing which must be carried out at a location relatively close to the lens output. The front-stage processing includes black-level balancing processing, noise elimination processing and shading correction processing.

The front-stage processing section 121 carries out the black-level balancing processing, the noise elimination processing and the shading correction processing on the long exposure-time image.

On the other hand, the front-stage processing section 122 carries out the black-level balancing processing, the noise elimination processing and the shading correction processing on the short exposure-time image.

The amplifier 123 multiplies the long exposure-time image already subjected to the front-stage processing carried out in the front-stage processing section 121 by a gain received from the CPU 131 in order to produce a product and supplies the product to the synthesizing processing section 126.

By the same token, the amplifier 124 multiplies the short exposure-time image already subjected to the front-stage processing carried out in the front-stage processing section 122 by the gain received from the CPU 131 in order to produce a product and supplies the product to the synthesizing processing section 126.

The gain output from the CPU 131 can be the reciprocal of an exposure ratio or a corrected value of the reciprocal. The exposure ratio is a ratio of the shutter speed for the normally taken image to the shutter speed for the short exposure-time image.

On the other hand, the long exposure-time image and the short exposure-time image which have been subjected to the front-stage processing are supplied to the color-flicker detection processing section 125.

The color-flicker detection processing section 125 examines the data of the long exposure-time image and the short exposure-time image in order to determine whether or not color flickers exist in the data. Then, the color-flicker detection processing section 125 also computes the degree of certainty at which the color flickers exist in the data and supplies information on the existence/nonexistence of the color flickers as well as the degree of certainty to the CPU 131. A method adopted by the color-flicker detection processing section 125 to detect the existence of the color flickers will be explained later in detail.

The synthesizing processing section 126 carries out processing to synthesize the long exposure-time image and the short exposure-time image in pixel units.

As a pixel synthesizing method, the synthesizing processing section 126 carries out processing to compare the pixel value of a pixel in the long exposure-time image with the pixel value of the corresponding pixel in the short exposure-time image in order to select one of the two pixels or to blend both the pixels. As an alternative pixel synthesizing method, the synthesizing processing section 126 carries out processing to add an offset and obtain a sum of the pixel values of both the pixels.

The synthesizing processing section 126 outputs a synthesized image obtained as a result of synthesizing the long exposure-time image and the short exposure-time image as well as a flag for every pixel of the synthesized image. The flag for a pixel indicates the long exposure-time image or the short exposure-time image from which the pixel has been extracted.

The synthesizing processing section 126 outputs the synthesized image to the RAW-RGB conversion processing section 127 which then carries out RAW-RGB conversion processing on the synthesized image.

The RAW-RGB conversion processing is processing to convert RAW data output by the solid-state image taking device 110 for every pixel into three pieces of data, i.e., R data, G data and B data. The RAW-RGB conversion processing is also referred to as de-mosaic processing.

It is to be noted that the RAW data includes values of a Bayer array of R, G and B colors or complementary colors and pixel values each obtained for a pixel by passing the pixel value through a color filter and carrying out opto-electrical conversion on the pixel value.

The RAW-RGB conversion processing section 127 outputs the R, G and B synthesized images obtained as a result of the RAW-RGB conversion processing to the integrated-value computation processing section 128 and the white-balance adjustment processing section 129.

First of all, the following description explains the configurations of the white-balance adjustment processing section 129 and the rear-stage processing section 130 as well as functions carried out by the white-balance adjustment processing section 129 and the rear-stage processing section 130.

In accordance with the long/short exposure-time select flag received from the synthesizing processing section 126, the white-balance adjustment processing section 129 multiplies the pixel values of the R, G and B synthesized images received from the RAW-RGB conversion processing section 127 by their respective R, G and B white-balance gains received from the CPU 131 for pixels of the long exposure-time image and the short exposure-time image.

In the following description, the R, G and B white-balance gains received from the CPU 131 for pixels of the long exposure-time image and the short exposure-time image are referred to as WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short.

The rear-stage processing section 130 carries out rear-stage processing on an image received from the white-balance adjustment processing section 129. The rear-stage processing includes noise elimination, an edge emphasis process, gradation conversion and gamma processing.

As described above, the RAW-RGB conversion processing section 127 outputs the R, G and B synthesized images obtained as a result of the RAW-RGB conversion processing to the integrated-value computation processing section 128.

The integrated-value computation processing section 128 determines pixels each meeting a given condition on the basis of the R, G and B synthesized images and the long/short exposure-time select flag, carrying out processing to integrate the values of every pixel. Finally, the integrated-value computation processing section 128 outputs every integration result of R, G and B images obtained as a result of the processing carried out on pixels of the entire screen to the CPU 131.

The given condition mentioned above implies area specification, which is specification of position coordinates of an image, and specification of a range from a certain level of a pixel value to another certain level of the pixel value. In addition, the given condition may also imply a condition requiring the long/short exposure-time select flag to indicate that the long exposure-time image or the short exposure-time image has been selected.

Detailed Description of the Color-Flicker Detection

A method adopted by the color-flicker detection processing section 125 to detect color flickers is explained in detail as follows.

FIG. 5 shows a flowchart representing processing to detect color flickers in accordance with the first embodiment.

The flowchart begins with a step ST1 at which a RAW-RGB conversion process is carried out. Then, at the next step ST2, the color-flicker detection processing section 125 computes R, G and B integrated values for the long exposure-time image and the short exposure-time image. This computation of integrated values is similar to the integrated-value computation carried out by the integrated-value computation section 5 employed in the configurations shown in FIGS. 2 and 3.

As the main process of the integrated-value computation, the color-flicker detection processing section 125 integrates the values of pixels meeting certain conditions. Finally, the color-flicker detection processing section 125 outputs an integration result of the integrated-value computation carried out on pixels of the entire screen to the CPU 131E.

FIG. 6 shows a flowchart representing the processing to integrate R, G and B pixel values, computing the integration results at the step ST2. In the flowchart shown in FIG. 6, the color-flicker detection processing section 125 determines whether or not a pixel meets the certain conditions by determining whether or not the position of the pixel is in a specified area and determining whether or not the value of the pixel is in a specified range. For a pixel meeting these two conditions, RGB pixel value integration processing is carried out (refer to steps ST11 to ST16 of the flowchart shown in FIG. 6).

A method for specifying an integration range of pixel values is explained as follows.

FIG. 7 is a conceptual diagram referred to in the following description of a method for specifying a range of integration of pixel values for a case in which a noise margin is not taken into consideration. The figure shows relations between the sensor output representing the pixel value and the light quantity for the long exposure-time image and the short exposure-time image.

This embodiment is characterized in that an integration range is so specified that the integration range includes pixel values to be integrated for both the long exposure-time image and the short exposure-time image.

As is obvious from FIG. 7, for light quantities smaller than Pa, the output generated by the sensor to represent the pixel value for the long exposure-time image is lower than the saturation level. Thus, the pixel value generated by the sensor exists. For light quantities greater than Pa, on the other hand, the sensor gets saturated inevitably so that the sensor outputs no meaningful information.

Accordingly, it is nice to specify an integration range including pixel values produced by light quantities smaller than Pa for the long exposure-time image and short exposure-time image. In the typical example shown in the figure, pixel values included in a portion denoted by symbol X are pixel values to be integrated.

FIG. 7 is a conceptual diagram referred to in the above description of the method for specifying an integration range of pixel values for a case in which noises are not taken into consideration.

In an actual image taking operation, however, noises do exist and, the larger the exposure-time ratio, the smaller the number of pixels included in the short exposure-time image as integration-object pixels each generating a low-level signal. Thus, a large exposure-time ratio deteriorates the precision of the determination as to whether or not color flickers exist. In order to solve this problem, the integration range can be specified by taking a noise margin into consideration in advance.

FIG. 8 is a conceptual diagram to be referred to in the following description of a method for specifying a range of integration of pixel values for a case in which a noise margin is taken into consideration. In the diagram, a noise level is provided on the sensor-output axis at the sensor output of 0 and the sensor saturation level. As shown in FIG. 8, pixel values included in the noise margins are not subjected to integration.

Thus, pixel values to be integrated are sensor outputs in a portion denoted by symbol Y in the figure. In comparison with the case in which a noise margin is not taken into consideration, the number of pixel values to be integrated is small.

By specifying such a range of integration of pixel values and integrating the pixel values, six final integration results can be obtained at the step ST2 of the flowchart shown in FIG. 5. The final integration results are R, G and B integration results for the long exposure-time image and R, G and B integration results for the short exposure-time image. In the following description, the R, G and B integration results for the long exposure-time image and R, G and B integration results for the short exposure-time image are denoted by reference notations Sum_R_Long, Sum_G_Long, Sum_B_Long, Sum_R_Short, Sum_G_Short and Sum_B_Short respectively.

Then, by making use of Equations (1-1) to (1-4) given below as equations based on the computed integration results, R/G and B/G ratios for the long exposure-time image and the short exposure-time image are computed at a step ST3 of the flowchart shown in FIG. 5 as follows:
[Equations 1]
(R/G)_Long=SumR_Long/SumG_Long  (1-1)
(B/G)_Long=SumB_Long/SumG_Long  (1-2)
(R/G)_Short=SumR_Short/SumG_Short  (1-3)
(B/G)_Short=SumB_Short/SumG_Short  (1-4)

Typical ratios computed by making use of Equations (1-1) to (1-4) are shown in FIGS. 9 and 10 for the short exposure-time image and the long exposure-time image respectively.

FIG. 9 is a conceptual diagram showing a ratio of R to G and a ratio of B to G for a short exposure-time image.

FIG. 10 is a conceptual diagram showing a ratio of R to G and a ratio of B to G for a long exposure-time image.

Both the figures show typical ratios for a case in which color flickers are generated. The ratios shown in the figure are no more than typical ratios since the dent and protrusion of each graph representing the ratio and the upper-lower relation between the graphs vary from situation to situation. That is to say, the graph representing the R/G ratio is not always above the graph representing the B/G ratio and the graph representing the B/G ratio is not necessarily a graph having a dent.

As described before, the short exposure-time image is an image taken at a relatively high shutter speed. Thus, the short exposure-time image is easily affected by color flickers, and the R/G and B/G ratios change with the lapse of time as shown by the graphs of FIG. 9.

On the other hand, the long exposure-time image is an image taken at a relatively low shutter speed. Thus, the long exposure-time image is hardly affected by color flickers. However, the long exposure-time image may not be said to be an image not affected by color flickers at all. In comparison with curves representing the R/G and B/G ratios for the short exposure-time image, the R/G and B/G ratios for the long exposure-time image change with the lapse of time only a little bit as shown by the curves of FIG. 10.

Then, at the next step ST4, evaluation values of the color flickers are computed.

The evaluation values of the color flickers are computed in accordance with Equations (2-1) and (2-2) given as follows:
[Equations 2]
CR_EstR=(R/G)_Short/(R/G)_Long  (2-1)
CR_EstB=(B/G)_Short/(B/G)_Long  (2-2)

FIG. 11 is a diagram showing typical color-flicker evaluation values obtained by making use of the above equations.

As indicated by Equations (2-1) and (2-2), an evaluation value is a ratio for the short exposure-time image to a ratio for the long short exposure-time image. By examining the evaluation value, the flicker components having an effect also on the long exposure-time image can be canceled from the short exposure-time image. In addition, the ratio for the short exposure-time image is normalized automatically with respect to the ratio for the long exposure-time image.

Even though the short exposure-time image and the long exposure-time image have a difference in exposure time, each particular pixel on the long exposure-time image is compared with a pixel on the short exposure-time image at the same position as the particular pixel. Thus, information on image taking of the same image taking object should be obtained.

Accordingly, if no color flickers are generated, the short exposure-time image has B/G and R/G ratios equal to those of the long exposure-time image regardless of the exposure time. At that time, the values to be compared as described above are the B/G and R/G ratios. Thus, even if the magnitude of a pixel value (or the absolute pixel value) of the long exposure-time image is different from that of the short exposure-time image, no problem is raised. Accordingly, if no color flickers are generated, the evaluation values which are each a ratio should be 1.0.

If color flickers are generated, on the other hand, as described earlier, even for the same image-taking object, the hue changes due to the effect of the afterglow characteristic exhibited by the illuminated light. In this case, the evaluation values which are each a ratio each have a value greater or smaller than 1.0.

For the reason described above, if color flickers are generated, the evaluation values change from the value of 1.0 with the lapse of time as shown by curves of FIG. 11.

Then, at steps ST5 to ST8, the evaluation values are used to determine whether or not color flickers exist.

As explained before, in an ideal state with no color flickers generated, the evaluation values should be 1.0.

However, the following facts exist:

1): Errors caused by noises and the like exist.

2): Even though the flicker components having an effect on the long exposure-time light can be canceled from the short exposure-time light in accordance with Equations (5-1) and (5-2), the effect of the flickers components on the long exposure-time image is not zero.
3): An integrated value is not a result of integration of evaluation values but a result of integrating each of R, G and B pixel values.

Thus, there is a difference from a case in which an evaluation value is obtained from R, G and B integrated values. As a result, there is almost no case in which the evaluation value is just equal to 1.0.

For this reason, a dead zone NZN like one shown in FIG. 12 is defined and a color-flicker evaluation value existing in this dead zone NZN is regarded as a value indicating that no color flickers have been generated.

To be more specific, the evaluation value CR_Est_R for the R color and the evaluation value CR_Est_B for the B color are examined to determine whether or not the evaluation values CR_Est_R and CR_Est_B are in the dead zone NZN. If at least one of the evaluation values CR_Est_R and CR_Est_B is outside the dead zone NZN, the existence of color flickers is confirmed.

The above description has explained a method for determining whether or not color flickers exist.

If a result of the determination method indicates that color flickers exist, the existing color-flicker suppression method described before can be applied.

FIG. 13 is a block diagram showing a typical configuration of the color-flicker detection processing section 125.

As shown in the figure, the color-flicker detection processing section 125 employs an RAW-RGB conversion section 1251, a condition determination section 1252, an RGB integration section 1253, an evaluation-value computation section 1254 and a detection section 1255.

These sections carry out their respective pieces of processing in order to accomplish the processing to detect color flickers as described before.

Color-Flicker Suppression Method

As explained earlier, color flickers are detected in accordance with this embodiment and, if the existence of the color flickers is confirmed, the ordinary color-flicker suppression method can be applied.

On the other hand, additional processing can also be carried out besides the color-flicker detection according to the embodiment. The additional processing is carried out in order to compute the degree of certainty at which a result of the color-flicker detection is obtained. Then, it is possible to apply a color-flicker suppression method based on the degree of certainty.

The following description explains a method for computing the degree of certainty and a method for suppressing color flickers on the basis of the degree of certainty.

First of all, the following description explains the method for computing the degree of certainty at which a result of the color-flicker detection is obtained.

FIG. 14 shows a flowchart representing color-flicker detection and certainty-degree computation processing.

As shown in FIG. 14, the flowchart representing the processing begins with a step ST21 at which color flickers are detected. Color flickers are detected in the same way as the color-flicker detection processing represented by the flowchart shown in FIG. 6.

The result of the color-flicker detection carried out at the step ST21 is examined at the next step ST22 in order to determine whether or not color flickers exist. If color flickers do not exist, the flow of the processing represented by the flowchart shown in FIG. 14 goes on to a step ST23 in order to compute white-balance gains as multipliers common to long and short exposure-time images from R, G and B integrated values.

The R, G and B integrated values used in the computation of the white-balance gains can be integrated values computed at the flicker detection time or integrated values computed by the integrated-value computation section. One of differences between the integrated values computed at the flicker detection time and the integrated values computed by the integrated-value computation section shown is FIGS. 2 to 4 is conceivably a difference in conditions such as the computation range.

In addition, a method for computing white-balance gains as multipliers common to long and short exposure-time images from R, G and B integrated values can be a gain computation method based on Equations (3-1) to (3-3) given below. However, Equations (3-1) to (3-3) are no more than typical equations. That is to say, any gain computation method can be adopted as long as the gain computation method is an already existing method.
[Equations 3]
WBGR=(SumG_Long+SumG_Short)/(SumR_Long+SumR_Short)  (3-1)
WBGG=1.0  (3-2)
WBGB=(SumG_Long+SumG_Short)/(SumB_Long+SumB_Short)  (3-3)

Then, at the next ST24, the white-balance adjustment processing block carries out white-balance adjustment processing to multiply long and short exposure-time images by the common white-balance gains.

If color flickers exist, on the other hand, processing described below is carried out at a step ST25 in order to compute the degree of certainty at which the existence of the color flickers has been detected.

FIG. 15 is a diagram referred to a method for computing the degree of certainty at which color flickers are detected.

FIG. 16 is a diagram showing degrees of certainty at which color flickers are detected.

The degree of certainty is defined as the amount of deviation of a color-flicker evaluation value computed at a color-flicker detection time from a dead zone NZN as shown in FIG. 15. In the figure, the length of an arrow is the degree of certainty. In this case, the amount of deviation is handled as a scalar quantity. Thus, the degree of certainty is always a positive number. The certainty degree obtained by adoption of such a method is shown in FIG. 16.

FIG. 17 is a diagram referred to in the following explanation of another method for computing the degree of certainty at which color flickers are detected.

In accordance with the method explained above by referring to FIG. 15, the degree of certainty is defined as the amount of deviation of a color-flicker evaluation value from a dead zone NZN. As shown in FIG. 17, however, the degree of certainty is defined as the amount of deviation of a color-flicker evaluation value from the value of 1. The certainty degree defined as shown in FIG. 17 can also be obtained by setting each of TH_high and TH_low shown in FIG. 15 at 0.

Next, the following description explains a method adopted at a step ST26 to compute a white-balance gain, which is used for suppressing color flickers, on the basis of the computed degree of certainty.

At the step ST26, first of all, white-balance gains are computed as multipliers for each of the long exposure-time image and the short exposure-time image from the R, G and B integrated values.

As a method for computing the white-balance gains, a computation method based on Equations (4-1) to (4-6) given below can be adopted. Much like Equations (3-1) to (3-3), Equations (4-1) to (4-6) are also typical equations.
[Equations 4]
R_gain_Long=SumG_Long/SumR_Long  (4-1)
G_gain_Long=1.0  (4-2)
B_gain_Long=SumG_Long/SumB_Long  (4-3)
R_gain_Short=SumG_Short/SumR_Short  (4-4)
G_gain_Short=1.0  (4-5)
B_gain_Short=SumG_Short/SumB_Short  (4-6)

The R, G and B integrated values used in computing the white-balance gains in accordance with Equations (4-1) to (4-6) can be integrated values computed at the flicker detection time or integrated values computed by the integrated-value computation section shown in FIGS. 2 to 4. One of differences between the integrated values computed at the flicker detection time and the integrated values computed by the integrated-value computation section is conceivably a difference in conditions such as the computation range.

From the six white-balance gains R_gain_Long, G_gain_Long, B_gain_Long, R_gain_Short, G_gain_Short and B_gain_Short computed in accordance with Equations (4-1) to (4-6) respectively, six final white-balance gains WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short are obtained to be used eventually as multipliers for the long exposure-time image and the short exposure-time image.

FIG. 18 is a diagram showing a white-balance gain computation system 140 for computing the different final white-balance gains for short and long exposure-time images.

As shown in the figure, the white-balance gain computation system 140 employs multipliers 141 and 142 as well as an adder 143.

A blending ratio α is computed from a color-flicker certainty degree obtained at the step ST25. Then, the white-balance gains for the long exposure-time image are blended with the white-balance gains for the short exposure-time image on the basis of the blending ratio α.

FIG. 19 is an explanatory diagram referred to a method for computing the blending ratio α.

The horizontal axis of FIG. 19 represents the color-flicker certainty degree obtained at the step ST25 whereas the vertical axis of the figure represents the blending ratio α.

For certainty degrees smaller than a lower-side threshold value TH_min, the blending ratio α is set at 0. For certainty degrees greater than an upper-side threshold value TH_max, the blending ratio α is set at 1.0.

The blending ratio α for any particular certainty degree between the lower-side threshold value TH_min and the upper-side threshold value TH_max has the value in the range 0 to 1.0. The value in the range 0 to 1.0 is determined from a straight line connecting a point representing the blending ratio α of 0 and the lower-side threshold value TH_min to a point representing the blending ratio α of 1.0 and the upper-side threshold value TH_max as shown in FIG. 19. To put it concretely, the blending ratio α for the particular degree of certainty is obtained by multiplying the difference between the lower-side threshold value TH_min and the particular degree of certainty with the gradient of the straight line. The gradient of the straight line is equal to 1.0/(TH_max−THh_min).

By adoption of the method described above, it is possible to find the six final white-balance gains WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short to be used eventually as multipliers for the long exposure-time image and the short exposure-time image. Then, these multipliers are used by the white-balance adjustment processing section in order to adjust the white balance.

In accordance with the method explained above by referring to FIG. 18, the blending ratio α is used in a blending process carried out on the six white-balance gains R_gain_Long, G_gain_Long, B_gain_Long, R_gain_Short, G_gain_Short and B_gain_Short in order to compute the six final white-balance gains WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short. However, the blending process can also be carried out on the R, G and B integrated values in place of the six white-balance gains R_gain_Long, G_gain_Long, B_gain_Long, R_gain_Short, G_gain_Short and B_gain_Short in order to compute the six final white-balance gains WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short as shown in FIG. 20. As shown in the figure, the blending process is carried out on six R, G and B integrated values SUM_R_Long, SUM_G_Long, SUM_B_Long, SUM_R_Short, SUM_G_Short and SUM_B_Short in order to compute the six final white-balance gains WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short.

The above description explains a method for suppressing color flickers by carrying out color-flicker detection in accordance with this embodiment.

FIG. 21 is a block diagram showing a typical configuration of a color-flicker detection processing section 125A having a function to suppress color flickers.

As shown in the figure, the color-flicker detection processing section 125A employs a condition determination section 1252, an RGB integration section 1253, an evaluation-value computation section 1254, a detection section 1255, a certainty-degree computation section 1256 and a white-balance gain computation section 1257.

These sections carry out their respective pieces of processing in order to accomplish the processing to detect color flickers and suppress the flickers.

FIG. 22 is a block diagram showing a typical configuration of an image taking apparatus 100B employing an image processing apparatus according to the second embodiment of the present disclosure.

FIG. 23 shows a flowchart representing processing to detect color flickers in accordance with the second embodiment.

FIG. 24 is a block diagram showing a typical configuration of a color-flicker detection processing section 125B according to the second embodiment.

A difference between the image taking apparatus 100B according to the second embodiment and the image taking apparatus 100 according to the first embodiment is described as follows:

The first embodiment is an embodiment in which the white-balance adjustment processing is carried out at the rear stage in the same way as the configuration shown in FIG. 3.

On the other hand, the second embodiment is an embodiment in which the white-balance adjustment processing is carried out at the front stage in the same way as the configuration shown in FIG. 2.

Thus, in the case of the second embodiment, the RAW-RGB conversion is not required in the processing to detect color flickers as shown in FIGS. 23 and 24.

The third embodiment of the present disclosure is explained below to serve as an embodiment in which static/dynamic-state determination is added to the conditions applied to the R, G and B integrated values.

FIG. 25 shows a flowchart representing processing for a case in which static/dynamic-state determination for R, G and B integrated values is included as an additional condition in accordance with the third embodiment of the present disclosure.

Typical Configuration for Adding Static/Dynamic-State Determination to Conditions for RGB Integrated Values

In the case of the first embodiment, values of pixels having information meaningful for both the long time exposure image and the short time exposure image are integrated. For the same pixel, except for color flickers, it is assumed that the ratio for the long time exposure image is different from the ratio for the short time exposure image. Ideally, the evaluation value is 1.0.

Since there is a small difference in image taking time between the long exposure-time image and the short exposure-time image, however, a moving object may pass through during this short time difference. If a moving object passes through during this short time difference, the presumption of the same image taking object is undesirably no longer valid so that the result of color-flicker detection is affected.

In order to solve the problem described above, in the third embodiment, static/dynamic-state determination is carried out for every pixel or every block of pixels as a newly added condition referred to as a static-state condition. In this way, color flickers can be detected with a high degree of precision (refer to a flowchart shown in FIG. 25).

In the third embodiment, as is the case with the existing technology, a mechanism for obtaining a motion vector or information in the course of the process of obtaining the motion vector may be used in order to determine a static or dynamic state on the basis of data such as differences between preceding and succeeding frames along the time axis.

FIG. 26 is a block diagram showing a typical configuration of the color-flicker detection processing section 125C provided with a static/dynamic-state determination function in accordance with the third embodiment.

As shown in the figure, the color-flicker detection processing section 125C employs a condition determination section 1252, an RGB integration section 1253, an evaluation-value computation section 1254, a detection section 1255, a static/dynamic-state determination section 1258 and a memory used as a delay circuit 1259.

These sections carry out their respective pieces of processing in order to accomplish the processing to detect color flickers.

A typical embodiment described below as the fourth embodiment of the present disclosure integrates evaluation values instead of integrating R, G and B integrated values.

FIG. 27 shows a flowchart representing processing to integrate evaluation values in accordance with the fourth embodiment of the present disclosure instead of integrating R, G and B integrated values.

FIG. 28 shows a flowchart representing color-flicker determination processing for a case in which the processing to integrate evaluation values has been carried out.

Typical Configuration for Integrating Evaluation Values in Place of R, G and B Integrated Values

In the case of the embodiments described before, R, G and B integrated values are computed and a color-flicker evaluation value is obtained from the integrated values.

As explained earlier, however, there is a difference between a configuration in which an evaluation value is obtained from R, G and B integrated values and a configuration in which an evaluation value is obtained for every pixel and, then, an integrated value is obtained as the average of the evaluation values.

For the latter case, FIG. 27 is given as a figure showing a flowchart representing processing to integrate evaluation values at a step ST36 in accordance with the fourth embodiment whereas FIG. 28 is given as a figure showing a flowchart representing color-flicker determination processing for a case in which the processing to integrate evaluation values has been carried out as shown in the flowchart of FIG. 27.

That is to say, a process carried out to compute the average of evaluation values at a step ST41 of the flowchart shown in FIG. 28 is described in detail in the flowchart shown in FIG. 27.

The flowchart shown in FIG. 27 begins with a step ST31 at which the integrated values of the evaluation values for the color flickers are reset. Then, at the next step ST32, a pixel value is received.

Subsequently, at the next step ST33, the pixel is examined in order to determine whether or not the pixel is located at a position in a specified area. If the pixel is located at a position in the specified area, at a step ST34, the pixel value is examined in order to determine whether or not the pixel value is within a specified range. If the pixel value is within the specified range, at a step ST35, R/G and B/G ratios are computed for each of the long time exposure image and the short time exposure image whereas evaluation values for color flickers are obtained from the R/G and B/G ratios. Then, at the next step ST36, an evaluation-value sum representing a sum of evaluation values for color flickers is computed.

Subsequently, at the next step ST37, a pixel count representing the number of pixels processed so far is updated. Then, at the next step ST38, the present pixel is examined in order to determine whether or not the present pixel is the last processed pixel. If the present pixel is not the last processed pixel, the processing to integrate evaluation values is repeated. As a matter of fact, the processing to integrate evaluation values is carried out repeatedly till the last pixel is processed.

The flowchart shown in FIG. 28 begins with a step ST41 at which the evaluation-value sums obtained at the step ST36 of the flowchart shown FIG. 27 are divided by the pixel count obtained at the step ST37 of the flowchart shown FIG. 27 in order to find B/G and R/G average evaluation values.

Then, at the next step ST42, the B/G average evaluation value is examined in order to determine whether or not the B/G average evaluation value is in the range of a dead zone. If the B/G average evaluation value is in the range of the dead zone, the R/G average evaluation value is examined at a step ST43 in order to determine whether or not the R/G average evaluation value is in the range of the dead zone. If the R/G average evaluation value is in the range of the dead zone, nonexistence of color flickers is confirmed at a step ST44.

If the B/G average evaluation value is obtained at the step ST42 to be not in the range of the dead zone and/or if the R/G average evaluation value is obtained at the step ST43 to be not in the range of the dead zone, existence of color flickers is confirmed at a step ST45.

In a typical embodiment explained below as the fifth embodiment of the present disclosure, processing to divide a color-flicker detection area is carried out.

FIG. 29 serves as an explanatory diagram referred to a case in which processing to divide a color-flicker detection area is carried out in accordance with the fifth embodiment as well as a diagram showing a typical situation in which a plurality of optical sources exist.

FIG. 30 is an explanatory diagram referred to a case in which processing to divide a color-flicker detection area is carried out in accordance with the fifth embodiment.

Processing to Divide a Color-Flicker Detection Area

In accordance with the embodiments described earlier, more robust color-flicker detection is possible so that generated color flickers can be suppressed effectively.

In the case of the embodiments described earlier, however, each of the long time exposure image and the short time exposure image can be multiplied only by a set of white-balance gains, i.e., the R, G and B gains.

In consequence, if a plurality of optical sources with characteristics different from each other exist in the short time exposure image, only color flickers caused by any one of the optical sources can be dealt with.

In the case of a situation like the one shown in FIG. 29 for example, optical sources providing information meaningful for the short time exposure image are an indoor lamp serving as an optical source OS1 and a sunlight source serving as an optical source OS2 which is an external optical source that can be seen through a window. In such a situation, if color flickers caused by the indoor lamp are suppressed, white-balance adjustment can no longer be carried out on the external scene which can be seen through the window.

In order to solve the above problem, as shown in FIG. 30, the screen is divided into several areas DVA for each of which the same processing can be carried out.

If the screen is divided into several areas DVA, however; for an area DVA, the short time exposure image may not exist and the number of pixels may be too small so that correct values may not be integrated. In such a case, if the computed pixel count is too small, control is executed to cancel the suppression of color flickers.

FIG. 31 is a block diagram showing a typical configuration of an image taking apparatus 100C employing an image processing apparatus 120C according to the sixth embodiment of the present disclosure.

The image taking apparatus 100C according to the sixth embodiment is different from the image taking apparatus 100 according to the first embodiment in that, in the case of the image taking apparatus 100, the typical processing of the first embodiment is carried out by hardware. The typical processing includes the processing to compute an evaluation value.

In order to find a ratio in the processing to compute an evaluation value, however, a divider is required so that it is quite within the bounds of possibility that the size of the hardware circuit of the image taking apparatus 100 increases.

In order to solve the above problem, the sixth embodiment adopts a method to make use of the CPU 131C to carry out some of the processing. To put it concretely, in the case of the sixth embodiment, the color-flicker detection processing section 125C carries out processing up to the integration process by making use of hardware whereas the CPU 131C carries out the computation of an evaluation value and the determination as to whether or not color flickers exist.

FIG. 32 is a block diagram showing a typical configuration of an image taking apparatus 100D employing an image processing apparatus 120D according to a seventh embodiment of the present disclosure.

The image taking apparatus 100D according to the seventh embodiment is different from the image taking apparatus 100B according to the second embodiment in that, in the case of the image taking apparatus 100B, the typical processing of the second embodiment is carried out by hardware. The typical processing includes the processing to compute an evaluation value.

In order to find a ratio in the processing to compute an evaluation value, however, a divider is required so that it is quite within the bounds of possibility that the size of the hardware circuit of the image taking apparatus 100B increases.

In order to solve the above problem, the seventh embodiment adopts a method to make use of the CPU 131D to carry out some of the processing. To put it concretely, in the case of the seventh embodiment, the color-flicker detection processing section 125D carries out processing up to the integration process by making use of hardware whereas the CPU 131D carries out the computation of an evaluation value and the determination as to whether or not color flickers exist.

Each of eighth and ninth embodiments of the present disclosure implements a typical configuration in which delay adjustment is carried out in order to cope with delayed white-balance gains used for detecting and suppressing color flickers.

FIG. 33 is a block diagram showing a typical configuration of an image taking apparatus 100E employing an image processing apparatus 120E according to the eighth embodiment of the present disclosure.

The image taking apparatus 100E according to the eighth embodiment is different from the image taking apparatus 100 according to the first embodiment in that, in the case of the image taking apparatus 100E, delay adjustment is carried out in order to make a frame, from which evaluation values have been computed, the same as a frame for which white-balance gains have been computed for the white-balance adjustment processing. Thus, the image taking apparatus 100E employs additional memories 132 and 133 for the delay adjustment.

FIG. 34 is a block diagram showing a typical configuration of an image taking apparatus 100F employing an image processing apparatus 120F according to a ninth embodiment of the present disclosure.

The image taking apparatus 100F according to the ninth embodiment is different from the image taking apparatus 100B according to the second embodiment in that, in the case of the image taking apparatus 100F, delay adjustment is carried out in order to make a frame, from which evaluation values have been computed, the same as a frame for which white-balance gains have been computed for the white-balance adjustment processing. Thus, the image taking apparatus 100F employs additional memories 132 and 133 for the delay adjustment.

Delay Adjustment for Delayed White-Balance Gains Used for Detecting and Suppressing Color Flickers

In the typical processing carried out by the embodiments described earlier, white-balance gains for each of the long exposure-time image and the short exposure-time image are computed from integration results obtained from the integration processing carried out on all pixels of an entire frame. Thus, in order to complete the integration processing, it takes time corresponding to one frame so that the white-balance gains are delayed accordingly. In addition, the CPU 131 also carries out processing to compute evaluation values from the integration results so that it also takes time as well to compute the evaluation values.

Thus, in the typical processing carried out by the embodiments described earlier, a frame from which evaluation values have been computed by carrying out integration processing is different from a frame for which the white-balance adjustment is to be executed by making use of the white-balance gains in actuality.

In the case of the eighth and ninth embodiments, on the other hand, a frame from which evaluation values have been computed is made the same as a frame for which the white-balance adjustment is to be executed by making use of the white-balance gains. By adding the memories 132 and 133 to the configuration shown in FIGS. 4 and 22, delay adjustment can be carried out so as to make a frame, from which evaluation values have been computed, the same as a frame for which the white-balance adjustment is to be executed.

It is to be noted that characteristics peculiar to the embodiments described so far can be properly combined with each other.

In accordance with the first to ninth embodiments, the following effects can be obtained.

In comparison with the existing white-balance adjustment processing, more robust color flicker detection is possible.

White-balance adjustment processing proper for the degree of certainty at which color flickers are detected can be carried out so that the color flickers can be suppressed effectively.

Even for an image with a wide dynamic range, proper white-balance adjustment processing can be carried out so that the color flickers can be suppressed effectively.

Since determination of the existence of color flickers is carried out for each frame without taking the time-axis direction periodicity into consideration, the white-balance adjustment processing can be carried out on color flickers other than periodical color flickers. Thus, proper suppression of color flickers is possible.

In addition to the function to detect color flickers, a tenth embodiment of the present disclosure also has a function to suppress generation of a false color in a blend area of the long exposure-time image and the short exposure-time image.

FIG. 35 is a block diagram showing a typical configuration of an image taking apparatus employing an image processing apparatus according to the tenth embodiment of the present disclosure.

The image processing apparatus employs a synthesizing processing section 126G, a RAW-RGB conversion processing section 127G, an integrated-value computation processing section 128G, a white-balance adjustment processing section 129G and a CPU 131G.

These sections carry out their respective pieces of processing in order to accomplish the function to detect color flickers and the function to suppress generation of a false color in a blend area.

The synthesizing processing section 126G carries out processing to synthesize the long exposure-time image with the short exposure-time image in pixel units. Details of the synthesizing method will be described later.

As a result of the synthesizing processing, the synthesizing processing section 126G outputs a synthesized image and flag information and/or a blending ratio α of the synthesizing. The flag information is provided for each synthesized-image pixel of the synthesized image to indicate which of a pixel of the long exposure-time image and a pixel of the short exposure-time image has been selected as the synthesized-image pixel. In the following description, the flag information is referred to as a long/short exposure-time select flag FLGLS.

The synthesizing processing section 126G outputs the synthesized image to the RAW-RGB conversion processing section 127G which then carries out RAW-RGB conversion processing on the synthesized image.

The RAW-RGB conversion processing is processing to convert RAW data output by the solid-state image taking device 110 for every pixel into three pieces of data, i.e., R data, G data and B data. The RAW-RGB conversion processing is also referred to as de-mosaic processing.

It is to be noted that the RAW data includes values of a Bayer array of R, G and B colors or complementary colors and pixel values each obtained for a pixel by passing the pixel value through a color filter and carrying out opto-electrical conversion on the pixel value.

The RAW-RGB conversion processing section 127G outputs the R, G and B synthesized images obtained as a result of the RAW-RGB conversion processing to the integrated-value computation processing section 128G and the white-balance adjustment processing section 129G.

First of all, the following description explains the configurations of the white-balance adjustment processing section 129G and the rear-stage processing section 130 as well as functions carried out by the white-balance adjustment processing section 129G and the rear-stage processing section 130.

In accordance with the long/short exposure-time select flag FLGLS and blending ratio α received from the synthesizing processing section 126G, the white-balance adjustment processing section 129G multiplies the pixel values of the R, G and B synthesized images received from the RAW-RGB conversion processing section 127G by their respective R, G and B white-balance gains received from the CPU 131G for pixels of the long exposure-time image and the short exposure-time image.

In the following description, the R, G and B white-balance gains received from the CPU 131G for pixels of the long exposure-time image and the short exposure-time image are referred to as WBG_R_Long, WBG_G_Long, WBG_B_Long, WBG_R_Short, WBG_G_Short and WBG_B_Short.

As described earlier, the long/short exposure-time select flag FLGLS is a flag indicating which of a pixel of the long exposure-time image and a pixel of the short exposure-time image has been selected as the synthesized-image pixel.

The rear-stage processing section 130 carries out rear-stage processing on an image received from the white-balance adjustment processing section 129G as described above. The rear-stage processing includes noise elimination, an edge emphasis process, gradation conversion and gamma processing.

As described above, the RAW-RGB conversion processing section 127G outputs the R, G and B synthesized images obtained as a result of the RAW-RGB conversion processing to the integrated-value computation processing section 128G.

The integrated-value computation processing section 128G determines pixels each meeting a given condition on the basis of the R, G and B synthesized images and the long/short exposure-time select flag FLGLS, carrying out processing to integrate the values of every pixel. Finally, the integrated-value computation processing section 128G outputs every integration result of R, G and B images obtained as a result of the processing carried out on pixels of the entire screen to the CPU 131G.

The given condition mentioned above implies area specification, which is specification of position coordinates of an image, and specification of a range from a certain level of a pixel value to another certain level of the pixel value. In addition, the given condition may also imply a condition requiring the long/short exposure-time select flag FLGLS to indicate that the long exposure-time image or the short exposure-time image has been selected.

In the following description, the integration results produced by the integrated-value computation processing section 128G are referred to as Sum_R_Long, Sum_G_Long, Sum_B_Long, Sum_R_Short, Sum_G_Short and Sum_B_Short.

From the R, G and B integration results supplied by the integrated-value computation processing section 128G, the CPU 131G computes the white-balance gains WBG_R_Long, WBG_G_Long and WBG_B_Long for the long exposure-time image as well as the white-balance gains WBG_R_Short, WBG_G_Short and WBG_B_Short for the short exposure-time image.

As a method for computing the white-balance gains, the CPU 131G adopts a computation method like a typical one based on Equations (5-1) to (5-6) given below. However, the computation method based on Equations (5-1) to (5-6) is no more than a typical method or the computation method adopted by the CPU is not limited particularly to this typical method. That is to say, the CPU 131G may adopt any computation method as long as the method is an existing method according to a technology in related art.
[Equations 5]
WBGR_Long=SumG_Long/SumR_Long  (5-1)
WBGG_Long=1.0  (5-2)
WBGB_Long=SumG_Long/SumB_Long  (5-3)
WBGR_Short=SumG_Short/SumR_Short  (5-4)
WBGG_Short=1.0  (5-5)
WBGB_Short=SumG_Short/SumB_Short  (5-6)
Detailed Description of the Synthesizing Processing

The following description explains details of a synthesizing method for suppressing a false color generated in a blend area.

FIG. 36 is a block diagram showing a typical configuration of the synthesizing processing section 126G adopting a synthesizing method for suppressing a false color generated in a blend area in accordance with the tenth embodiment.

As shown in the figure, the synthesizing processing section 126G employs RAW-YC conversion processing sections 1261 and 1262, a selector 1263, a blending-ratio computation section 1264, multipliers 1265 and 1266, a subtractor 1267, an adder 1268 and a YC-RAW conversion processing section 1269.

The RAW-YC conversion processing section 1261 carries out RAW-YC conversion processing on the RAW data RAW_Long of the long exposure-time image received by the RAW-YC conversion processing section 1261.

The RAW-YC conversion processing section 1262 carries out RAW-YC conversion processing on the RAW data RAW_Short of the short exposure-time image received by the RAW-YC conversion processing section 1262.

The RAW-YC conversion processing is processing to convert RAW data into a luminance (Y) component and a color (C) component.

The conversion formula used in the RAW-YC conversion processing varies in accordance with the origin and the evaluation value of the RAW data, the luminance (Y) component and the color (C) component. Here, the origin of the RAW data indicates which color filter has been used by the image taking device to acquire image-taking information. The color filter can be an R (Red), G (Green), B (Blue), C (Cyan), M (Magenta), Y (Yellow) or G (Green) filter. The evaluation value indicates an existing color space which can be YCrCb, YUV or YPrPb.

Equations (6-1) to (6-3) given below are typical conversion formulas of transformation into YCrCb for a case in which the color filter of the image taking device is a filter of the RGB primary-color system.

Equations (7-1) to (7-3) given below are available as equations for expressing relations between the RGB primary-color system and the CMYG complementary-color system. Thus, by substituting Equations (7-1) to (7-3) into Equations (6-1) to (6-3), the CMYG complementary-color system can be transformed into YCrCb.
[Equations 6]
Y=0.257×R+0.504×G+0.098×B  (6-1)
Cr=−0.148×R−0.291×G+0.439×B  (6-2)
Cb=0.439×R−0.368×G−0.071×B  (6-3)
[Equations 7]
C=1.0−R  (7-1)
M=1.0−G  (7-2)
Y=1.0−B  (7-3)

The equations given above are typical transformation formulas. Other transformation formulas or simplified formulas can also be used. In addition, other evaluation values can also be used as well. On top of that, there are no particular restrictions on how to make use of pixel information or surrounding-pixel information in the conversion processing described above.

As results of the conversion processing, the RAW-YC conversion processing sections 1261 and 1262 generate luminance (Y) components and color (C) components for the long exposure-time image and the short exposure-time image respectively. The luminance (Y) component is subjected to synthesizing processing making use of a blending ratio mα also referred to as a mixing ratio in the blending-ratio computation section 1264, the multipliers 1265 and 1266 as well as the subtractor 1267.

In the synthesizing processing, the blending-ratio computation section 1264 computes the blending ratio mα by making use of the luminance (Y) component of the long exposure-time image and the luminance (Y) component of the short exposure-time image. The method for computing the blending ratio mα can be the existing technique.

The color (C) component is not subjected to the synthesizing processing. The color (C) component of the long exposure-time image or the color (C) component of the short exposure-time image is selected by the selector 1263 in accordance with a C-component select signal CSEL and the selected color (C) component is used by the YC-RAW conversion processing section 1269.

The selected one of the color (C) component of the long exposure-time image and the color (C) component of the short exposure-time image does not raise a problem. In the case of a system for generating an image having a wide dynamic range, however, the signal level of the short exposure-time image is low. Thus, it is quite within the bounds of possibility that a relatively large number of noises are generated in the short exposure-time image.

For the reason described above, it is quite within the bounds of possibility that a selected color (C) component of the long exposure-time image is better than a selected color (C) component of the short exposure-time image in many cases.

Finally, the YC-RAW conversion processing section 1269 makes use of the blended luminance (Y) component and the selected color (C) component to carry out processing of converting the components back into RAW data. The processing is based on a conversion method making use of formulas obtained by inverting the formulas described earlier.

As explained before, the synthesizing processing section 126G outputs a synthesized image obtained as a result of the synthesizing processing and a long/short exposure-time select flag FLGLS provided for each synthesized-image pixel of the synthesized image to indicate which of a pixel of the long exposure-time image and a pixel of the short exposure-time image has been selected as the synthesized-image pixel.

Here, in the case of the tenth embodiment, the value of the long/short exposure-time select flag FLGLS for the blend area indicates that the color (C) component of the long exposure-time image or the color (C) component of the short exposure-time image has been selected. Thus, a method for determining the value of the long/short exposure-time select flag FLGLS can be represented by a flowchart shown in FIG. 36.

FIG. 37 shows a flowchart representing processing to determine the value of a long/short exposure-time select flag FLGLS in accordance with the tenth embodiment.

The flowchart representing the above processing begins with a step ST51 at which a pixel value is received. Then, at the next step ST52, a blending ratio mα is computed. If the computed value of the blending ratio mα is obtained at the next step ST53 to be 0.0, the ID (identifier) of the short exposure-time image is output at a step ST54. If the computed value of the blending ratio mα is obtained at the next step ST53 to be a value in the range of 0.0<mα<1.0, the ID (identifier) of an image having the color (C) component thereof selected is output at a step ST55. If the computed value of the blending ratio mα is obtained at the next step ST53 to be 1.0, the ID (identifier) of the long exposure-time image is output at a step ST56.

In either case, the flow of the processing then goes on to the next step ST57 in order to determine whether or not the last pixel has been processed. The processing is carried out repeatedly till the last pixel has been processed.

The synthesizing processing section 126G according to the tenth embodiment has been described above.

FIGS. 38A and 38B are conceptual diagrams referred to in the following explanation of a reason why a false color generated in a blend area can be suppressed in accordance with the tenth embodiment.

FIG. 38A is a conceptual diagram referred to in the following explanation of a synthesizing processing method according to the tenth embodiment.

In the tenth embodiment, the color (C) component is not subjected to synthesizing processing. That is to say, the color (C) component of an image taken at a long or short exposure time is used as it is. Thus, also in the white-balance adjustment processing section 129G provided at the rear stage, the used color (C) component is just multiplied by a white-balance gain for the color (C) component.

As a result, as shown in FIG. 38B, in a blend area PXMX, a proper white-balance gain can be used as a multiplier to suppress the false color generated in the blend area PXMX.

The following description explains eleventh to fifteenth embodiments which mainly carry out processing to suppress generation of a false color in a blend area of the long exposure-time image and the short exposure-time image.

FIG. 39 is a block diagram showing a typical configuration of an image taking apparatus 100H employing an image taking apparatus 120H according to the eleventh embodiment of the present disclosure.

The image taking apparatus 100H according to the eleventh embodiment is different from the image taking apparatus 100G according to the tenth embodiment in that, in the case of the eleventh embodiment, a new signal is transmitted from the synthesizing processing section 126H to the white-balance adjustment processing section 129H to convey a blending ratio. The synthesizing processing section 126H and the white-balance adjustment processing section 129H will be explained in detail later.

Thus, in addition to the synthesized-image signal and the long/short exposure-time select flag FLGLS, the synthesizing processing section 126H in the eleventh embodiment also outputs the blending ratio (BLD ratio) ma used in the synthesizing processing to the white-balance adjustment processing section 129H.

Since the other sections are identical with the corresponding ones employed in the tenth embodiment, their explanation is not repeated.

Detailed Description of the Synthesizing Processing

The following description explains details of a synthesizing method for suppressing a false color generated in the blend area in accordance with the eleventh embodiment.

FIG. 40 is a diagram showing a typical configuration of the synthesizing processing section 126H applying the synthesizing method to taken images in order to suppress the false color generated in the blend area in accordance with the eleventh embodiment. Components composing the synthesizing processing section 126H are explained one after another as follows.

As shown in FIG. 40, the synthesizing processing section 126H employs a blending-ratio computation section 1264H, multipliers 1265H and 1266H, a subtractor 1267H and an adder 1268H.

First of all, the synthesizing processing section 126H carries out synthesizing processing on the RAW data RAW_Long of the long exposure-time image and the RAW data RAW_Short of the short exposure-time image.

In the synthesizing processing, the blending-ratio computation section 1264H computes the blending ratio mα by making use of the RAW data RAW_Long of the long exposure-time image and the RAW data RAW_Short of the short exposure-time image. As the method for computing the blending ratio, the blending-ratio computation section 1264H may adopt a method to compute the blending ratio in accordance with pixel values or a method adopted by a modification version to be explained later.

The RAW data is blended by the multipliers 1265H and 1266H on the basis of the computed blending ratio mα which is also output by the synthesizing processing section 126H to the white-balance adjustment processing section 129H provided at the rear stage.

In addition, the method adopted by the synthesizing processing section 126H to determine the value of the long/short exposure-time select flag FLGLS which is one of the outputs of the synthesizing processing section 126H is the same as the method explained earlier by referring to the flowchart shown in FIG. 37 in the description of the tenth embodiment.

The above description explains the synthesizing processing section 126H according to the eleventh embodiment.

Detailed Description of the White-Balance Adjustment Processing

The following description explains details of a white-balance adjustment method adopted by the white-balance adjustment processing section 129H according to the eleventh embodiment to suppress a false color generated in a blend area.

FIG. 41 is a diagram showing a typical configuration of a white-balance adjustment processing section 129H applying the synthesizing method to white-balance gains in order to suppress a false color generated in a blend area in accordance with the eleventh embodiment.

As shown in FIG. 41, the white-balance adjustment processing section 129H employs multipliers 1291 to 1293, a subtractor 1294 and an adder 1295.

In accordance with an existing technology, the white-balance adjustment processing is carried out by making use of the long/short exposure-time select flag FLGLS as shown in FIG. 4.

In the white-balance adjustment processing section 129H according to the eleventh embodiment, on the other hand, the long/short exposure-time select flag FLGLS is not used. Instead, the white-balance adjustment processing is carried out by making use of a blending ratio mα received from the synthesizing processing section 126H for every pixel as a ratio used in synthesizing processing. The sequence of the white-balance adjustment processing is described as follows.

Signals input to the white-balance adjustment processing section 129H include R, G and B images obtained as a result of RAW-RGB conversion and the blending ratio mα used in the synthesizing processing. In addition, input signals supplied to the white-balance adjustment processing section 129H also include the white-balance gains WBG_R_Long, WBG_G_Long and WBG_B_Long for the long exposure-time image as well as the white-balance gains WBG_R_Short, WBG_G_Short and WBG_B_Short for the short exposure-time image. These white-balance gains have been computed by the CPU 131 on the basis of RGB integrated values.

On the basis of the blending ratio mα used in the synthesizing processing, the circuit of the white-balance adjustment processing section 129H shown in FIG. 41 computes final white-balance gains WBG_R, WBG_G and WBG_B for every pixel from the white-balance gains WBG_R_Long, WBG_G_Long and WBG_B_Long for the long exposure-time image as well as the white-balance gains WBG_R_Short, WBG_G_Short and WBG_B_Short for the short exposure-time image.

Then, white-balance adjustment processing is carried out on the R, G and B images by making use of the final white-balance gains WBG_R, WBG_G and WBG_B.

The white-balance adjustment processing section 129H according to the eleventh embodiment has been explained above.

In the case of the existing white-balance adjustment processing section, pixels of one screen serve as a multiplicand whereas two groups of multipliers are used for the multiplicand. The first group has the white-balance gains WBG_R_Long, WBG_G_Long and WBG_B_Long for the long exposure-time image whereas the second group has the white-balance gains WBG_R_Short, WBG_G_Short and WBG_B_Short for the short exposure-time image.

In the case of the eleventh embodiment, on the other hand, the white-balance adjustment processing section 129H is utilized in order to make it possible to make use of white-balance gains, which each vary from pixel to pixel, as a multiplier. As a result, combinations of a number of white-balance gains are used as multipliers for pixels on one screen.

The above description explains the synthesizing processing section 126H and the white-balance adjustment processing section 129H which are included in the eleventh embodiment.

FIGS. 42A and 42B are conceptual diagrams of explaining a reason why a false color generated in a blend area can be suppressed by making use of the synthesizing processing section 126H and the white-balance adjustment processing section 129H which are provided in accordance with the eleventh embodiment.

FIG. 42A is a conceptual diagram of synthesizing processing carried out in accordance with the eleventh embodiment.

In accordance with the eleventh embodiment, the synthesizing processing section 126H carries out the synthesizing processing on the RAW data of the long exposure-time image and the RAW data of the short exposure-time image in the same way as the existing technology.

The white-balance adjustment processing section 129H according to the eleventh embodiment multiplies the value of every pixel in the blend area, in which the RAW data of the long exposure-time image has been blended with the RAW data of the short exposure-time image, by a white-balance gain obtained by the CPU 131 for the pixel. In this way, for every pixel, it is possible to carry out white-balance adjustment based on the blending ratio used in the synthesizing processing.

As a result, it is possible to solve the problem of the technique in related art by substantially suppressing the false color generated in the blend area as shown in FIG. 42B.

In an embodiment described below as the twelfth embodiment of the present disclosure, processing to compute the blending ratio mα used in the synthesizing processing includes static/dynamic-state determination.

FIG. 43 shows a flowchart representing processing to compute the blending ratio in accordance with the twelfth embodiment of the present disclosure.

Typical Processing to Compute the Blending Ratio mα by Making Use of Static/Dynamic-State Determination

The method explained before by referring to FIG. 7 is adopted by the tenth and eleventh embodiments as a method for computing the blending ratio mα.

Since there is a small difference in image taking time between the long exposure-time image and the short exposure-time image, however, a moving object may pass through during this short time difference. If a moving object passes through during this short time difference, the moving object probably exists only on the long exposure-time image or the short exposure-time image in some cases.

In such cases, if the synthesizing processing is carried out by making use of a blending ratio obtained in accordance with the level of the pixel, as a result, information on pixels of entirely different image-taking objects is blended. Thus, the reader may easily imagine a case in which the synthesizing processing for generating a synthesized image ends in a failure.

In order to solve the problem described above, in the case of the twelfth embodiment, the flowchart shown in FIG. 43 includes a step ST62 executed to determine whether a pixel or a pixel area also referred to as a pixel block is a static one pertaining to an image of an image taking object or a dynamic one pertaining to an image of a moving object.

If the pixel or the pixel area is determined to be a static one, the value of the blending ratio mα is computed at a step ST63.

If the pixel or the pixel area is determined to be a dynamic one, on the other hand, the blending processing is not carried out. Instead, the pixel value or the luminance level is compared with a threshold value LV_TH at a step ST64. In accordance with the result of the comparison, the short exposure-time image is selected at a step ST65, or the long exposure-time image is selected at a step ST66.

The processing represented by this flowchart can also be carried out in the first and eleventh embodiments.

FIG. 44 is a block diagram showing a typical configuration of a synthesizing processing section 1261 provided with a static/dynamic-state determination function according to the twelfth embodiment to serve as a modification of the synthesizing processing section 126 shown in FIG. 36.

As shown in FIG. 44, the synthesizing processing section 126I employs RAW-YC conversion processing sections 1261I and 1262I, a selector 1263, a blending-ratio computation section 1264I, multipliers 1265I and 1266I, a subtractor 1267I, an adder 1268I, a YC-RAW conversion processing section 1269I and a static/dynamic-state determination section 1270.

These sections carry out their respective pieces of processing in order to accomplish the synthesizing processing described above.

FIG. 45 is a block diagram showing a typical configuration of a synthesizing processing section 126J provided with a static/dynamic-state determination function according to the twelfth embodiment to serve as a modification of the synthesizing processing section 126H shown in FIG. 40.

As shown in FIG. 45, the synthesizing processing section 126J employs a blending-ratio computation section 1264J, multipliers 1265J and 1266J, a subtractor 1267J, an adder 1268J and a static/dynamic-state determination section 1270J.

These sections carry out their respective pieces of processing in order to accomplish the synthesizing processing described above.

Each of thirteenth and fourteenth embodiments of the present disclosure implements a typical configuration in which the blending ratio mα is computed and delay adjustment is carried out in order to cope with the delayed white-balance gains for suppression processing.

FIG. 46 is a block diagram showing a typical configuration of an image taking apparatus 100B employing an image processing apparatus 120B according to the thirteenth embodiment of the present disclosure.

The image taking apparatus 100K according to the thirteenth embodiment is different from the image taking apparatus 100 according to the tenth embodiment in that the image taking apparatus 100K employs an additional memory 132 for the delay adjustment. The delay adjustment is carried out in order to make a frame, from which evaluation values have been computed, the same as a frame for which white-balance gains have been computed for the white-balance adjustment processing.

The fourteenth embodiment of the present disclosure implements a typical configuration in which the blending ratio mα is computed and delay adjustment is carried out in order to cope with delayed white-balance gains for suppression processing.

FIG. 47 is a block diagram showing a typical configuration of an image taking apparatus 100L employing an image processing apparatus 120L according to the fourteenth embodiment of the present disclosure.

The image taking apparatus 100L according to the fourteenth embodiment is different from the image taking apparatus 100H according to the eleventh embodiment in that the image taking apparatus 100L employs an additional memory 132 for delay adjustment. The delay adjustment is carried out in order to make a frame, from which evaluation values have been computed, the same as a frame for which white-balance gains have been computed for the white-balance adjustment processing.

Delay Adjustment for Blending Ratio mα and White-Balance Gains for Suppression Processing

In the tenth and eleventh embodiments, the white-balance gains for each of the long exposure-time image and the short exposure-time image are computed from integration results by adoption of the ordinary technique. In accordance with the ordinary technique, in the integration processing carried out to produce the integration results by the integrated-value computation processing section 128 of the configurations shown in FIGS. 35 and 39, pixels of the entire screen are used.

Accordingly, in order to complete the integration processing, it takes time longer than a time period corresponding to at least one frame. As a result, the computation of the white-balance gains in the CPU 131 are delayed.

In addition, the CPU 131 also carries out processing to compute evaluation values from the integration values so that it also takes time as well to compute the evaluation values. Thus, in the case of the tenth and eleventh embodiments, a frame from which evaluation values have been computed from the integration results obtained by carrying out the integration processing is different from a frame for which white-balance gains are computed for the white-balance adjustment processing. That is to say, a frame from which evaluation values have been computed is different from a frame for which the white-balance adjustment is to be executed by making use of the white-balance gains in actuality.

By making use of the circuits according to the tenth and eleventh embodiments as well as the processing method adopted by these embodiments, white-balance adjustment proper for every individual pixel can be carried out in comparison with the ordinary technique. Due to this frame difference, it is quite within the bounds of possibility that the performance to suppress the false color deteriorates.

In the case of the thirteenth and fourteenth embodiments, on the other hand, the memory 132 is added for delay adjustment in order to make a frame, from which evaluation values have been computed, the same as a frame for which white-balance gains have been computed for the white-balance adjustment processing.

By adding the memory 132 to each of the image taking apparatus 100G shown in FIG. 35 and the image taking apparatus 100H shown in FIG. 39 to obtain respectively image taking apparatus 100K shown in FIG. 46 and the image taking apparatus 100L shown in FIG. 47 as described above, it is possible to carry out delay adjustment in the image taking apparatus 100K and the image taking apparatus 100L.

In the fifteenth embodiment of the present disclosure, the long/short exposure-time select flag is omitted.

FIG. 48 is a block diagram showing a typical configuration of an image taking apparatus 100M employing an image processing apparatus 120M according to the fifteenth embodiment of the present disclosure.

The image taking apparatus 100M according to the fifteenth embodiment is different from the image taking apparatus 100H according to the eleventh embodiment in that, in the case of the image taking apparatus 100M according to the fifteenth embodiment, the long/short exposure-time select flag is omitted.

Typical Configuration without the Long/Short Exposure Time Select Flag

The long/short exposure-time select flag is a signal indicating pixels of which of the long exposure-time image and the short exposure-time image have been selected in the synthesizing processing.

On the other hand, the value of the blending ratio mα proposed in the eleventh embodiment can be used to indicate that the pixels of the long exposure-time image only or that the pixels of the short exposure-time image only have been selected.

To put it concretely, mα=1.0 indicates that the pixels of the long exposure-time image have been selected whereas mα=0.0 indicates that the pixels of the short exposure-time image have been selected.

As is obvious from the above description, in the case of the eleventh embodiment, the long/short exposure-time select flag can be replaced by the blending ratio mα. Thus, it is not necessary to make use of the long/short exposure-time select flag.

For the reason described above, a typical configuration of an image taking apparatus 100M provided by the fifteenth embodiment shown in FIG. 48 is obtained by eliminating the long/short exposure-time select flag from the configuration of the eleventh embodiment shown in FIG. 39.

Merits of the Tenth and Eleventh Embodiments

Merits of the tenth and eleventh embodiments are described as follows.

Merits of the Tenth Embodiment

Since the processing according to the tenth embodiment is all but completed in the synthesizing processing section 126, a delay circuit (or the like) for delaying a signal supplied to the rear stage does not exist. Thus, the size of the hardware can be made small.

It is possible to make use of the color (C) component of an image signal generated by light of a long exposure time as a signal having few noises in comparison with an image signal generated by light of a short exposure time if the color (C) component of the long exposure-time image has been selected.

Merits of the Eleventh Embodiment

Since the white-balance gain varying from pixel to pixel can be used as a multiplier, in comparison with the tenth embodiment, proper suppression of a false color can be carried out.

Since RAW data is blended, there is no spatial changeover point between the long exposure-time image and the short exposure-time image as long as the color (C) component is concerned. That is to say, a smooth changeover can be made. Thus, it is difficult for the user to recognize a false color even if sufficient suppression of the false color cannot be carried out.

As described above, in accordance with the tenth to fifteenth embodiments, the following effects are exhibited.

In processing to generate an image with a wide dynamic range on the basis of a plurality of images taken with exposure times different from each other, a false color generated in a blend area of the taken images can be suppressed effectively. The false color is contained in the blend area and raises a problem because the optical source varies from taken image to taken image, causing color flickers (color rolling) to be generated.

Further, in comparison with the existing white-balance adjustment processing, more diversified white-balance adjustment processing can be carried out. That is to say, white-balance adjustment processing proper for an application can be carried out.

The image taking apparatus according to the embodiments described above can each be used as an apparatus having a variety of image taking functions. Typical examples of the apparatus having a variety of image taking functions are a portable phone, a video camera and a digital still camera.

It is to be noted that the method explained above in detail can be configured by implementing the method as a program which is to be executed by a computer such as a CPU to serve as a program according to the procedure of the method.

In addition, the program can be configured as a program stored in advance on a recording medium to be accessed by the computer in order to have the program executed by the computer on which the recording medium is mounted. Typical examples of the recording medium are a semiconductor memory, a magnetic disk, an optical disk and a floppy (a registered trademark) disk.

The present disclosure contains subject matter related to that disclosed in Japanese Priority Patent Application JP 2010-144207 filed in the Japan Patent Office on Jun. 24, 2010, the entire content of which is hereby incorporated by reference.

It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and alternations may occur depending on design requirements and other factors insofar as they are within the scope of the appended claims or the equivalent thereof.

Yamamoto, Hiroshi, Shinmei, Katsuhisa, Nagataki, Shingo, Wajima, Kenji

Patent Priority Assignee Title
8908064, Sep 27 2010 FUJIFILM Corporation Imaging apparatus and imaging method
9148583, Jul 19 2013 Fujitsu Limited Flicker detection method and flicker detection apparatus
9167173, Mar 15 2013 Canon Kabushiki Kaisha Image capturing apparatus configured to perform a cyclic shooting operation and method of controlling the same
Patent Priority Assignee Title
5966174, May 27 1994 Matsushita Electric Industrial Co., Ltd. Signal processing circuit and method for driving a solid state imaging device using plural reading operations per image scanning time
20080212956,
20090135276,
JP11075109,
JP2008085736,
JP7322147,
/////
Executed onAssignorAssigneeConveyanceFrameReelDoc
Jun 01 2011SHINMEI, KATSUHISASony CorporationASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS 0264760660 pdf
Jun 01 2011NAGATAKI, SHINGOSony CorporationASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS 0264760660 pdf
Jun 01 2011WAJIMA, KENJISony CorporationASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS 0264760660 pdf
Jun 02 2011YAMAMOTO, HIROSHISony CorporationASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS 0264760660 pdf
Jun 07 2011Sony Corporation(assignment on the face of the patent)
Date Maintenance Fee Events
Jun 27 2014ASPN: Payor Number Assigned.
Oct 23 2017REM: Maintenance Fee Reminder Mailed.
Apr 09 2018EXP: Patent Expired for Failure to Pay Maintenance Fees.


Date Maintenance Schedule
Mar 11 20174 years fee payment window open
Sep 11 20176 months grace period start (w surcharge)
Mar 11 2018patent expiry (for year 4)
Mar 11 20202 years to revive unintentionally abandoned end. (for year 4)
Mar 11 20218 years fee payment window open
Sep 11 20216 months grace period start (w surcharge)
Mar 11 2022patent expiry (for year 8)
Mar 11 20242 years to revive unintentionally abandoned end. (for year 8)
Mar 11 202512 years fee payment window open
Sep 11 20256 months grace period start (w surcharge)
Mar 11 2026patent expiry (for year 12)
Mar 11 20282 years to revive unintentionally abandoned end. (for year 12)